THE STRIKING OF PROOF AND PATTERN COINS IN THE EIGHTEENTH CENTURY G. P. DYER AND P. P. GASPAR THIS paper records the results of a study of proof, pattern, and currency pieces of the early milled period and of an examination of the surviving tools and dies from that period held by the Royal Mint. These two lines of inquiry began separately, but as the work progressed their inter-relationship became clear and the two finally merged, enabling us not only to formulate objective criteria for the identification of proof strikings but also to postulate a method of production for proof and pattern pieces different from that employed for ordinary coins. In this period proof and pattern coins are, of course, very often readily distinguish- able from their currency counterparts. Patterns by definition present few problems of identification, while on proofs the details of the design are normally more sharply defined and the field is generally superior as a result of the more careful polishing of the surface of the dies. Moreover, in the case of proofs of gold and silver coins, the edge is likely to be plain instead of grained or lettered. Closer examination, however, shows that these visible differences on patterns and proofs can be supplemented by others, less obvious perhaps but arguably more significant for the light they throw on the method of production. The most important of these additional differences relates to the lettering. Attention has often been drawn to the fact that on currency pieces letters with uprights have bases that are frequently fishtailed or bifurcated, and indeed this type of lettering is regarded as one of the most characteristic features of the eighteenth-century coinage.1 By contrast, on proofs and patterns the letter bases are usually square and unindented and seldom show the exaggerated serifs associated with currency pieces. This same firmness of outline is present also in the beading around the rim, for on proofs and patterns the beads are normally well defined and complete, whereas on currency pieces they are elongated and tend to disappear off the edge of the coin. The differences are well illus- trated by a pair of George II crowns of 1746 in the Royal Mint collection (PL V, 1). This example has been carefully selected to illustrate the distinction, and it would be as well to emphasize that there is not a hard and fast rule, for fishtailed letters can some- times be seen on proofs and square letters on currency pieces. The tendency, however, is undeniable. Another consistent feature of these early patterns and proofs is the evidence they bear of having received more than one blow from the dies. This normally shows itself in a partial doubling of the letters, but on occasion the displacement is much greater, as on a proof guinea of 1729 (PL V, 2). Doubling of busts and shields can also occur and, somewhat strangely, the doubling may be present on only one side of the coin. It is 1 W. J. Hocking, JVC, 4th ser. ix (1908), 97. 118 THE STRIKING OF PROOF AND PATTERN COINS easiest to see where a coin has been allowed to tone naturally, since the squashed letters from the previous blows have a tendency to tone less quickly and therefore show up as bright areas. To the square letter bases and the doubling can be added a cross-section that may frequently be wedge-shaped, a feature that is often accentuated by vertical burrs thrown up in striking. A very sharp burr may be seen, for example, on a Una and the Lion five-pound piece in the British Museum, and the authors have recently examined an 1831 twopence on which the burr on the obverse is so sharp and prominent that it can only be described as grotesque. These are nineteenth-century examples, and it should perhaps be stressed that the reference in the title of the paper to the eighteenth century is merely a convenient shorthand. The period to which these comments relate begins with the milled coinage and some of the features which have been observed are still apparent as late as the middle of the nineteenth century. The edges of proof and pattern coins often appear roughly filed, and in addition occasionally show a single step or raised witness line at some point around the circum- ference. The step is clearly shown on a pattern farthing of Queen Anne (PI. V, 3), where it forms a distinct and irregular ledge. By the late eighteenth century the step is no longer seen but has been replaced by a regular raised witness line, as on a proof Irish halfpenny of 1775 (PL V, 4). Finally, late eighteenth-century proofs usually have an upright die-axis and show a disregard for standard weight and fineness; they are generally made of fine gold or silver and their weights may vary substantially from standard and, for coins of the same denomination and date, from each other. It seemed that many of these differences could not be explained merely by more careful treatment or by heavier pressure in striking. On the contrary, the differences appeared to be related to each other and to reflect a fundamental difference in the method of striking. For this reason attention was increasingly focused on the surviving dies and the surviving mint records. The Royal Mint has a collection of over 10,000 obsolete master tools and dies, and although nineteenth- and twentieth-century material makes up the greater part of the collection there are some 600 or so items from the early milled period. Of these items more than half are punches, the remainder being matrices and dies. They were catalogued by an unknown hand about 18402 and were recatalogued and published by W. J. Hocking in 1910 as part of Volume II ofhis Royal Mint Museum Catalogue,3 In checking the tools against Hocking's catalogue, an immediate difficulty was encountered. This concerned a group of about twenty items with lettering and designs complete which Hocking had described as matrices,4 that is to say, the tools used to raise punches which in turn were used to sink dies. These tools do indeed resemble modern matrices by having gently tapering sides instead of a neck. Hocking assumed that they could not be dies, which by his time always had a neck in order to permit the use of a collar, and he was probably encouraged in this view by the presence in the collection of about seventy contemporary dies which do have necks. Thus he might 2 Anonymous manuscript catalogue of the die collec- vol. ii, Dies, Medals and Seals (1910). tion of the Royal Mint (c.l840) in the Royal Mint 4 Ibid., nos. 18, 19, 94, 99, 237, 324, 326, 353, 356, 360, Library, 1328 entries, 100 pp. 366,373,390,393-6,399,401,404,406, and 408. See P. P. 3 W. J. Hocking, Catalogue of the Coins, Tokens, Gaspar, BNJxWi (1977), 55. Medals, Dies and Seals in the Museum of the Royal Mint, IN THE EIGHTEENTH CENTURY 119 well have believed that dies with necks were normal while the much smaller number of items without necks served the more exotic purpose of raising punches. The identification of these items as matrices, however, raises impossible difficulties, because without exception all are beaded and lettered and no details of the designs are omitted. Hocking would have us believe, for example, that the quarter-guinea reverse tool of 1718 (PI. VI, 5), complete in all its details, is a matrix from which punches were made. If Hocking were right, this would mean that on all dies the position of the inscription in relation to the design would be consistent, but study of early milled coins indicates that there is no such consistency, that it is in fact the variation in the positioning of the inscription that is far and away the easiest means of distinguishing individual dies. Moreover, among the large number of punches that survive from this period none exists with beading and lettering, and on most of the reverse tools certain details of the design in addition to the inscription have been omitted and left for the engraver to add by hand to each individual die. Thus the quarter-guinea reverse punch (PI. VI, 6) is incomplete in all these ways and it is not easy to see how it could have been made from the complete and finished tool shown in PI. VI, 5. If anything, the opposite is true, for the other tool must surely be a die sunk from such a punch and then completed by hand. To be more specific, two of Hocking's so-called matrices are for the obverse and reverse of the Cromwell Crown of 1658.5 The obverse carries across the base of the effigy the same crack that is so evident on the coins and it is well known that the sur- viving coins show the development of this crack across the effigy. Now if this tool were a matrix, one would have to suppose either that a series of punches was raised from a progressively cracking matrix or that there was a series of matrices which all cracked in the same place but to varying extents. Either supposition is frankly absurd, and it is evident that the tool is the actual die that struck the coins. This conclusion persuaded us that all twenty or so items in this category were probably dies but caution dictated that independent means should be sought for confirmation. Fortunately, die records survive and indeed for the eighteenth century they are surprisingly complete.6 Under the terms of the Mint Indentures the engravers were expected at intervals to produce for inspection all the coinage tools in their possession.